U.S. patent number 8,161,829 [Application Number 12/844,999] was granted by the patent office on 2012-04-24 for circuit compensation in strain gage based transducers.
This patent grant is currently assigned to Vishay Precision Group, Inc.. Invention is credited to Thomas P. Kieffer, Robert B. Watson, Felix Zandman.
United States Patent |
8,161,829 |
Zandman , et al. |
April 24, 2012 |
Circuit compensation in strain gage based transducers
Abstract
A methodology for selecting and properly placing foil strain
gages on a transducer in a Wheatstone bridge, which provides a more
consistent creep response, especially when the transducer
temperature is changed. A transducer includes a counterforce
subjected to a predetermined physical load that provides tension
and compression strains (positive and negative, respectively). The
transducer also includes a plurality of strain gage grids that are
operatively attached to the counterforce in the tension and
compression strain areas of the counterforce and generate
electrical signals. The plurality of strain gages are electrically
connected in a Wheatstone bridge circuit where their electrical
signals due to creep are cancelled.
Inventors: |
Zandman; Felix (Malvern,
PA), Watson; Robert B. (Clayton, NC), Kieffer; Thomas
P. (Wake Forest, NC) |
Assignee: |
Vishay Precision Group, Inc.
(Malvern, PA)
|
Family
ID: |
42983668 |
Appl.
No.: |
12/844,999 |
Filed: |
July 28, 2010 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20110023630 A1 |
Feb 3, 2011 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
61229123 |
Jul 28, 2009 |
|
|
|
|
Current U.S.
Class: |
73/862.622;
73/862.627; 73/765 |
Current CPC
Class: |
G01L
1/2243 (20130101); G01L 1/2262 (20130101); G01L
1/2281 (20130101); G01B 21/045 (20130101); G01B
7/18 (20130101); Y10T 29/49007 (20150115) |
Current International
Class: |
G01L
1/00 (20060101); G01L 5/24 (20060101) |
Field of
Search: |
;73/766,862.623,862.627 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
2000-275116 |
|
Oct 2000 |
|
JP |
|
2003-322571 |
|
Nov 2003 |
|
JP |
|
Other References
Hoffmann, Karl, "Anwendung der Wheatstoneschen Bruckenschaltung".
cited by other .
International Search Report and Written Opinion for PCT Application
No. PCT/US10/043520 filed on Jul. 28, 2010. cited by other.
|
Primary Examiner: Caputo; Lisa
Assistant Examiner: Dunlap; Jonathan
Attorney, Agent or Firm: Volpe and Koenig, P.C.
Parent Case Text
CROSS REFERENCE TO RELATED APPLICATION(S)
This application claims the benefit of U.S. Provisional Application
No. 61/229,123 filed Jul. 28, 2009, which is incorporated by
reference as if fully set forth.
Claims
What is claimed is:
1. A transducer comprising: a counterforce subjected to a
predetermined physical load that provides tension and compression
strains (positive and negative, respectively); a plurality of
strain gage grids that are operatively attached to said
counterforce in the tension and compression strain areas of said
counterforce and generate electrical signals; and, said plurality
of strain gages are electrically connected in a Wheatstone bridge
circuit, wherein the electrical creep signals are substantially
equal in magnitude and of the same sign between tension and
compression strain areas, and when the equal electrical creep
signals from the tension and compression strain areas are combined
in the Wheatstone bridge circuit, the electrical creep signals are
electrically cancelled.
2. A temperature compensated transducer comprising: a counterforce
subjected to a predetermined physical load that provides tension
and compression strains that are positive and negative,
respectively; a plurality of strain gage grids that are operatively
attached to said counterforce in said tension and compression
strain areas and compensate for the physical changes in the
counterforce at a predetermined temperature T; and, said plurality
of strain gages generate electrical signals through the temperature
range that is T+/-200.degree. C.; and said plurality of strain
gages are electrically connected in a Wheatstone bridge circuit,
wherein the electrical creep signals are substantially equal in
magnitude and of the same sign between tension and compression
strain areas, and when the equal electrical creep signals from the
tension and compression strain areas are combined in the Wheatstone
bridge circuit, the electrical creep signals are electrically
cancelled throughout the temperature range T+/-200.degree. C.
3. A temperature compensated transducer comprising: a counterforce
that support a predetermined physical load that provides tension
and compression strains that are positive and negative,
respectively; a plurality of strain gage grids, operatively
attached to said counterforce in said tension and compression
strain areas in compensation for the physical changes in the
counterforce at a predetermined temperature T, said plurality of
strain gages generate electrical signals, that include a creep
component, through a temperature range that includes T; and, said
plurality of strain gages are electrically connected in a
Wheatstone bridge circuit, wherein the electrical creep signals are
substantially equal in magnitude and of the same sign between
tension and compression strain areas, and when the equal electrical
creep signals from the tension and compression strain areas are
combined in the Wheatstone bridge circuit, the electrical creep
signals are electrically cancelled throughout the temperature
range.
4. A method of manufacturing a transducer, the method comprising:
forming a transducer having a counterforce subjected to a
predetermined physical load that provides tension and compression
strains (positive and negative, respectively); providing a
plurality of strain gage grids that are operatively attached to the
counterforce in the tension and compression strain areas of the
counterforce and generate electrical signals; and connecting the
plurality of strain gages in a Wheatstone bridge circuit, wherein
the electrical creep signals are substantially equal in magnitude
and of the same sign between tension and compression strain areas,
and when the equal electrical creep signals from the tension and
compression strain areas are combined in the Wheatstone bridge
circuit, the electrical creep signals are electrically
cancelled.
5. The method of claim 4, wherein the plurality of strain gage
grids that are operatively attached to the counterforce in the
tension and compression strain areas and compensate for the
physical changes in the counterforce at a predetermined temperature
T, and the plurality of strain gages generate electrical signals
through the temperature range that is T+/-200.degree. C.
Description
FIELD OF INVENTION
The present disclosure relates to strain gages and more
particularly, to strain gages including creep compensation.
BACKGROUND
Strain gage based transducers are used in a variety of applications
to convert mechanical inputs (for example, weight, force, mass,
torque, pressure, deflection/displacement) into an electrical
output. The basis for all such devices is the same. Specifically, a
mechanical reaction device (commonly called a spring or
counterforce) is designed to respond to the specific input,
transducing the input into a measurable surface strain, which
changes proportionally with the applied input. Strain gages
attached to the transducer counterforce sense and respond to this
surface strain with a change in electrical resistance. The
counterforce is normally machined from high-quality tool steel
(e.g., 4340 or 4140), or highly processed (hardened/heat treated)
stainless steel (e.g., 17-4 PH or 17-7 PH), or high-grade, heat
treated aluminum (e.g., 2024-T351 or 2024-T81), or other excellent
spring materials like beryllium copper or N-Span C. However, there
are special cases where polymers are used (e.g., epoxy-glass
laminate, or cast/injection molded plastics), and where ceramic
materials are used (e.g., Al.sub.2O.sub.3 99+ percent). In fact,
over the course of transducer history, practically every
conceivable material has been used at one time or another as the
basis for a counterforce. The present invention is not limited to
any one material or even to a class of materials; it works well
with any material selected for use as a counterforce.
In all cases, strain gage based transducers are used to convert
physical loads or inputs into electrical outputs. Achieving the
highest level of transducer accuracy requires compensating the
device for certain accuracy-limiting effects; some of which are
inherent to the strain gage/transducer system, like creep, and some
of which are external effects, like changes in temperature, and
some of which are a combination, like creep change with
temperature, called creep TC. As an example, load cells are used in
the weighing industry as transducers to convert a weight
(mass/force) into a proportional electrical signal. The load cell
is designed mechanically to provide repeatable and
quasi-equal-magnitude surface strains at specific points, whereby
two of the strains are tensile (positive) and two are compressive
(negative). Electrical resistance strain gages bonded at these
points convert the surface strains resulting from an applied weight
into a proportional electrical signal. The strain gages are
connected into an electrical circuit, typically a Wheatstone
bridge, which optimizes the output signal.
In the Wheatstone bridge electrical circuit typically used in
transducers, four strain gages, plus a power source, are wired
together in the series/parallel circuit as depicted in FIG. 2. The
electrical nature of this circuit is such that when the bridge is
resistively balanced (i.e., all four gages are at nearly the same
resistance value) there is no voltage present across the output
terminals (O1 and O2). Conversely, when the strain gages are at
meaningfully different resistances, there can be a small voltage
measured across O1 and O2, proportional to the applied voltage,
V.sub.i. Specifically, when gages 1 and 3 increase in electrical
resistance and gages 2 and 4 simultaneously decrease in electrical
resistance, the maximum proportional output voltage is presented
across terminals O1 and O2. It is for this reason that transducer
designs incorporate positive and negative strains, so gages bonded
at those locations will increase and decrease resistance,
respectively, with applied weight; thus, maximizing the voltage
signal from the transducer for a given applied weight (maximizing
sensitivity).
Within the weighing industry there is a class of load cells used in
applications called legal-for-trade. These legal-for-trade load
cells must pass stringent qualification tests from internationally
recognized standards, such as OIML R60 (Organization Internationale
de Metrologie Legal). Results from these tests classify the load
cell over a specified temperature range (normally -10 to
+40.degree. C.) based upon achievable resolution of weight. The
classification metric used is divisions of resolution. For example,
a load cell having a maximum combined error of 0.033% is classified
as 3000D (3000 divisions) accuracy.
Several factors conspire to affect the classification category of a
load cell, including the mechanical design and production of the
load cell body, and performance characteristics of the strain gage
and its installation. Among the strain gage performance parameters,
creep is critical to load cell classification. Ignoring all other
error contributions, the allowable cord-slope creep within the
example classification (3000D) is 0.0233% FS/min. (percent
full-scale per minute).
Transducer creep is defined as a changing output with a stable
physical condition or input (weight, in the case of the load cell
example) under steady state environmental conditions. Strain gages
are custom designed to compensate for the inherent material creep
of specific transducer designs. A representative plot of creep for
the load cell example is shown in FIG. 3a, which also indicates the
chord slope value normally used to quantify the creep, even though
the figure clearly shows that creep is a nonlinear phenomenon.
Further, creep performance can change, and usually does change
using prior creep correction methods, when the transducer
temperature is changed from that which was used for initial creep
compensation (normally room temperature, T which is
.about.24.degree. C.). Changes in creep with temperature can
significantly affect the possible classification of a
legal-for-trade load cell. The results from creep measurements over
a specified temperature range are termed creep TC. In some server
application, the temperature range can be T+/-200.degree. C.
Several variables affect strain gage creep, including but not
limited to, the resistive material (electrical conductor) from
which the strain gage is produced, geometry (e.g., gage length,
cross-section dimensions, end loop size, shape, and orientation),
construction (materials used in building the gage, including
insulating backing and insulating overlay, if present), and
installation (thickness and type of cement, gage positioning). The
most common type strain gage used in transducers is the thin,
metal-foil variety, depicted schematically in FIG. 4. The gage
consists of a primary measurement length (gage length), L, a
primary measurement width (gage width), Z, a plurality of grid
lines, T, configured into a serpentine grid, R, with solder pad
connections, M, a plurality of creep controlling end loops, K, an
upper alignment guide, P, and a lower alignment guide, N, defining
the major measurement axis, J, and an insulating backing (carrier),
U.
Prior methods allow for convenient control of transducer creep at
room temperature to about 0.0175%/min of rated full-scale output;
or, when calculating from OIML R60 for the load cell example, a
little over 4000D. One prior method of achieving creep compensation
is to select the strain gage end loops (K in FIG. 4) to optimize
the creep component. This, of course, presumes proper control of
the other previously mentioned effects on creep. This method
utilizes four identical, or nearly identical strain gages, with the
end loop lengths chosen to provide a chord slope creep as small as
possible or, at the least, sufficient for the intended
classification. When attempting to achieve the lowest creep slope
possible (highest transducer resolution) from a production run of
transducers using this prior method, it is typically necessary to
grade the production lot, whereby all transducers from the lot are
tested and classified by their test results, with no a priori
guarantee that any individual transducer from the lot will achieve
a high standard.
A subtle variation on the above mentioned prior method of creep
compensation is to pick end loop lengths for the strain gages
slightly different from one another. With this method, there may be
three strain gages with equal end loop lengths and one different;
or, two gages with equal end loop lengths and the other two equal,
but different from the first two; or, all gages may have a slightly
different end loop length. This minor difference of creeping
characteristic is achieved using what might be referred to as
nearly identical strain gages. This practice primarily evolved from
the practical concern over what gages happened to be on-hand when
building the transducer, and happen to combine for a low creep
result; that is, the method evolved naturally because of inventory
practicality. While achieving an excellent creep result at one
temperature is possible using the method, it does not, however,
necessarily provide any improvement in creep TC performance over
the more commonly practiced use of identical strain gages.
Another method of achieving transducer creep compensation has been
suggested, whereby the overall stiffness of the strain gage is
altered by varying the amount of reinforcing fibers mixed with the
backing resin. This method is grounded in the relationship between
creep and the relative stiffness difference between the
counterforce and the strain gage. One obvious limitation with this
technique is its applicability only to mixed-resin backing systems,
which is not the dominant type used within the industry.
Achieving high resolution creep compensation over the entire -10 to
+40.degree. C. temperature range is a challenging aspect of these
prior methods. In another method, various electrical configurations
are designed into the strain gage circuit and are formed with the
strain gage grids at the time of etching. These configurations are
initially electrically inert, but when subsequently introduced into
the circuit as active elements by cutting appropriate electrical
shunts, the transducer can be creep compensated, including any
variation in creeping caused by a change in temperature. This work
is performed after the gage has been installed on the transducer.
Disadvantages of this method are 1) more complex and costly strain
gage design and production; and, 2) careful and selective
`trimming` of the creep characteristic in situ.
It is known that strain gage creep, as exhibited by transducer
output, is a viscoelastic phenomenon, as illustrated in FIG. 1. As
such, when utilizing prior methods of creep compensation, including
the method of choosing gages with identical end loops and the
method of choosing gages with nearly identical end loops, a typical
transducer creep result can be represented by the graph shown in
FIG. 1'. Shown in FIG. 5 is a corresponding representative
independent strain gage creep from the tension and compression
strain gages bonded to a load cell, which combine in the Wheatstone
bridge circuit to cause the total transducer creep. From FIG. 5, it
is obvious that the direction of creep, as represented by the two
curves for change in output with time (one curve for tension strain
gages and one for compression strain gages), is opposite in sign;
the tension gages are shown decreasing in output (becoming more
negative) and the compression gages are shown increasing in output
(becoming more positive). As noted previously for the
characteristic nature of the Wheatstone bridge, the electrical
result from the bridge (the proportional output voltage) of these
two opposing creep directions is an increase in that part of the
electrical signal caused by creep; subtracting opposite sign
signals results in addition of the two signals.
Thus, the prior methods have embraced a common result, where creep
compensation is achieved via physical cancelling (viz., the
positive counterforce creep is countered by the combined negative
tension/compression strain gage creep), but it has not addressed
the problem through electrical cancelling as disclosed herein.
OVERVIEW OF THE INVENTION
Disclosed herein is new methodology for selecting and properly
placing foil strain gages on a transducer in the Wheatstone bridge,
which provides a more consistent creep response, especially when
the transducer temperature is changed. An especially advantageous
transducer counterforce is the so called binocular or
reverse-bending design, whereby two locations on the counterforce
have small areas of concentrated tension strain and two locations
have small areas of concentrated compression strain.
Coincidentally, there is one each of the two opposite strains on
the top and one each of the two opposite strains on the bottom of
the counterforce. Strain gages with the appropriate creeping
characteristic, as disclosed herein, are attached to the
counterforce at these four locations; tension strain gages attached
over the tension areas and compression strain gages attached over
the compression areas. This counterforce design is particularly
attractive because effects from temperature gradients along the
length of the counterforce, not necessarily associated with
creeping, are naturally compensated by the bridge circuit.
Prior methods for correction of transducer creep relied on
precisely matching the inherent positive creep of deadweight loaded
transducers to an equal negative creep by the strain gages; thus, a
balance is struck between the two independent creeps [transducer
creep+(combined creep of the four strain gages)=0], resulting in a
steady transducer output with time. Intrinsic to this correction is
that total creep from the four strain gages is matched to
compensate the transducer creep, without concern for relative creep
matching between the strain gages (all gages are chosen equal or
nearly equal). By default, this requires that the combined strain
gage creep be opposite in sign to the transducer creep.
The techniques disclosed herein use the electrical nature of the
Wheatstone bridge circuit. Referring to FIG. 3a, instead of
matching strain gage creep precisely to the creeping of the
transducer counterforce, the two tension strain gages are picked to
provide an equal magnitude and same-sign creep result as the two
compression strain gages (both positive, for example). These
representative strain gage creep curves are shown in FIG. 6.
Resulting from the arithmetic of Wheatstone bridge circuits, when
the equal resistance changes caused by creep from the tension and
compression strain gages combine in the bridge circuit, they
cancel; thereby, cancelling the creeping result, as depicted in
FIG. 7. Thus, the combined strain gage creep is no longer an exact,
but opposite in sign match to the creep of the deadweight loaded
transducer counterforce, as in prior methods, but is instead a
match between the tension and compression strain gage creeping
characteristics, such that each will combine and cancel in the
Wheatstone bridge.
For example, it is only necessary to choose compression strain
gages with significant positive creeping, and then match the
tension strain gages with equal positive creeping. There is no
further need to pick strain gages with creeping precisely matched,
but opposite in sign, to the transducer counterforce.
An advantageous embodiment of this structure is stable transducer
creep performance when the transducer temperature is changed from
that at which initial creep correction is accomplished (improved
creep TC); for example, over the entire legal-for-trade temperature
range of -10 to +40.degree. C. When choosing tension strain gages
with matching positive creep to the compression strain gages [by
choosing tension strain gages with much shorter end loops than
those used for compression strain gages--one-half the length (50%),
for example], then the creep change with temperature can be made
equal between tension and compression strain gages and the creep
signal component at any temperature is cancelled in the Wheatstone
bridge. This advantageous embodiment is made possible because of
the similar creep TC characteristic obtained when strain gages are
properly selected for positive tension and compression creeping, as
opposed to the selection of identical or nearly identical strain
gages using prior methodology, where the tension and compression
strain gages exhibit different creep TC characteristics.
BRIEF DESCRIPTION OF THE DRAWINGS
A more detailed understanding may be had from the following
description, given by way of example in conjunction with the
accompanying drawings:
FIG. 1 illustrates that strain gage creep, as exhibited by
transducer output, is a viscoelastic phenomenon;
FIG. 1' is a graph showing typical transducer creep;
FIG. 2 shows a Wheatstone bridge electrical circuit typically used
in association with transducers;
FIG. 3a is a graph showing a representative plot of creep for a
load cell;
FIG. 3b is a graph showing physical creep cancellation action;
FIG. 4 is a drawing of a strain gage;
FIG. 5 is graph showing strain gage creep from tension and
compression strain gages bonded to a load cell (negative and
positive creep);
FIG. 6 is graph showing strain gage creep from tension and
compression strain gages bonded to a load cell (both negative
creep);
FIG. 7 is graph showing strain gage creep with creep
cancellation;
FIG. 8 is a drawing of a binocular transducer with counterforce
design;
FIG. 9a is a drawing of a binocular transducer showing the location
of the various electrical connections (1, 2, 3, 4);
FIG. 9b is Wheatstone bridge electrical circuit coupled to the
transducer of FIG. 9a.
FIG. 10 is a graph showing a plot of transducer creep (including
room temperature creep);
FIG. 11 is a graph showing a plot of transducer creep when a using
strain gages with identical end loops; and
FIG. 12 is a graph showing a plot of transducer creep when
combining physical and electrical cancellation.
DETAILED EXAMPLES
FIG. 8 illustrates a binocular transducer counterforce design. The
counterforce is fixed at one end, A, and loaded with a deadweight
(physical load) on the opposite end, B. The locations indicated by
C are areas of tension strain (positive) and the locations
indicated by D are areas of compressive strain (negative). Strain
gages are normally attached with their major measurement axis (J in
FIG. 4) aligned with the major measurement axis E of the
counterforce. If it is desired to compensate for other responses of
the transducer from weight or environment, the strain gage may be
offset slightly (rotated or displaced) from perfect alignment in
the direction of E, the disclosed structure remains valid and works
well.
The strain gages are fixed to the counterforce, such as by an
adhesive, direct lamination or some other process, and are wired
into a Wheatstone bridge circuit as illustrated in FIG. 9b. The
only criterion being that the four strain gages are suitably
electrically connected into a Wheatstone bridge circuit. Electrical
connections L.sub.1, L.sub.2, L.sub.3 and L.sub.4 form the supply
and signal terminal sets, respectively, for the transducer (FIG.
9a). Also shown in FIG. 9b is the governing equation for relative
bridge voltage output, e.sub.0/E, expressing the dependence of this
parameter on the strain levels of each strain gage. From this
expression, it can be demonstrated that the combined contribution
of strain gages 1 and 2, relative to bridge voltage output, is
subtractive. As such, if the same signal is present from both
gages, the combined result on relative bridge voltage is null (the
difference between two equal signals equals zero). This is the
basis for the structure disclosed herein. That being, choose strain
gages with matching creeping characteristic between tension and
compression gages, regardless of the transducer creeping
characteristic, and combine these two equal signals in the bridge
such that they cancel electrically.
In the prior methods, creep compensation is effected by the
physical relaxation of the strain gages countering the physical
extension of the counterforce. Since the strain gages relax at the
same rate as the counterforce is extending, there is no change in
electrical resistance of the strain gages and no change in output
signal from the Wheatstone bridge. This physical creep cancellation
action is depicted in FIG. 3b, where the creep curve for the strain
gages (lower curve), when subtracted from the creep curve for the
counterforce (upper curve), yields a stable electrical output
signal from the transducer.
In contrast to prior methods of physical creep correction, this
disclosure is directed to the use of the summing characteristic of
the Wheatstone bridge to achieve creep correction electrically. By
adjusting the strain gage creeping characteristic such that the
creep strain signal is equal in magnitude and of the same sign
between the tension and compression strain gages, then when these
equal creep signals are combined in the bridge (for strain gages 1
and 2), and for strain gages 3 and 4), the resultant is zero and
the creep signal is electrically cancelled.
As previously indicated, an advantage of this matching of creep
strain signal, and an important improvement is that by correctly
accomplishing the strain gage creep matching, creep TC is also
improved. In one example, the matching is accomplished by using a
very long end loop configuration for the compression strain gages,
relative to the tension strain gages. Specifically, when the web
thickness, F, in FIG. 8 is 2.10 mm, and the counterforce is
machined from 2024-T351 Aluminum, or equivalent (producing a
transducer with approximately 30 kg capacity), then using tension
strain gage end loops of 0.1778 mm length and compression strain
gage end loops of 0.4382 mm length, with all other design and
construction parameters between the two sets of strain gages being
equal, will provide excellent room temperature creep compensation
and creep TC, as compared to prior methods. It should be
appreciated that other methods of adjusting strain gage creep (by
changing relative grid design--lines and spaces--of the tension and
compression strain gages, for example; or, by changing carrier
material between the tension and compression gages, as another
example) are available and those methods are also claimed as part
of the present invention.
Using the above example, a plot of the resulting transducer creep
TC, which by default includes room temperature creep, is shown in
FIG. 10. In FIG. 11 is a plot of the equivalent transducer creep TC
when a prior method of creep correction is used; specifically,
strain gages with identical end loops. The improvement achieved is
demonstrated by a much flatter creep TC curve, thus providing
equivalent creep compensation at all three test temperatures, and,
thus, allowing a higher legal-for-trade transducer
classification.
The present invention has the additional benefit that is compatible
with and can be combined in the same transducer with the prior
methods to almost completely eliminate creep TC. Specifically,
prior methods utilize strain gages so that the combined
tension/compression strain gage creep cancels the physical creep of
the transducer element and the present invention selects strain
gages so that the positive and negative creeping characteristic
cancel each other. By combining physical and electrical
cancellation, the variation of creep with temperature is almost
completely eliminated as illustrated in FIG. 12.
It will be understood by those of skill in the art that the present
invention cancels only the variable component know generally as
creep and that the desired measurement the transducer is not
cancelled.
* * * * *